ggen0488-cyber/worldcup-2026-predictor-skill
Developer toolsThis page was translated automatically. Technical details may differ from the original. View the original
This project is used to predict after the AI Agent collects and saves schedule/result data and pre-match intelligence, combining editable Elo-style team ratings, a Poisson score model that takes football randomness by default, and Monte Carlo full-season tournament simulations. It also includes an optional metaphysical fun mode.
Overview
This project is used to predict after the AI Agent collects and saves schedule/result data and pre-match intelligence, combining editable Elo-style team ratings, a Poisson score model that takes football randomness by default, and Monte Carlo full-season tournament simulations. It also includes an optional metaphysical fun mode. This project is not an official FIFA project and is not affiliated, endorsed, or sponsored by FIFA. Please do not use the FIFA logo, official visuals, official fonts, or other protected brand assets in this project. This block is used to directly view the latest reference scores on the GitHub repository homepage. Below are the four group stage predictions for 2026-06-22 Beijing time, including completed scores, pre-match intelligence, and the model's default single-match randomness. Prediction results are for reference only; please watch the matches rationally. worldcup-2026-predictor/ is a portable skill directory. You can copy it to a compatible AI skill runtime environment or run scripts directly on the command line. Requires Python 3.9+, scripts do not rely on third-party libraries. During installation, only copy the worldcup-2026-predictor/ directory; data/live/ is the runtime cache and does not require committing or distributing. You can also install it as a project-level skill, copying the directory to the current project's .agents/skills/: In Codex, you can explicitly call $worldcup-2026-predictor, or directly request something like "Predict Argentina vs Brazil" for the Agent to match automatically. If no new skill appears, restart the Codex session. In Claude Code, you can call /worldcup-2026-predictor or trigger it in natural language. Claude Code reads SKILL.md and associated resources from the skill directory.
Overview (machine translated)
Chinese | English
An open-source, platform-neutral AI skill for predicting the probability of the 2026 World Cup.
This project is used to predict after the AI Agent collects and saves schedule/result data and pre-match intelligence, combining editable Elo-style team ratings, a Poisson score model that takes football randomness by default, and Monte Carlo full-season tournament simulations. It also includes an optional metaphysical fun mode.
This project is not an official FIFA project and is not affiliated, endorsed, or sponsored by FIFA. Please do not use the FIFA logo, official visuals, official fonts, or other protected brand assets in this project.
Latest predicted scores on GitHub homepage
This block is used to directly view the latest reference scores on the GitHub repository homepage. Below are the four group stage predictions for 2026-06-22 Beijing time, including completed scores, pre-match intelligence, and the model’s default single-match randomness. Prediction results are for reference only; please watch the matches rationally.
| Match | Maximum Probability Result | Main Probability | Summary of Predicted Reasons |
|---|---|---|---|
| Spain vs Saudi Arabia | Spain slightly favored, predicted score 2-0 | Spain win 77.1% / draw 14.7% / Saudi Arabia win 8.1% | Spain has clear advantages in rating and possession creativity, with expected goals around 2.41-0.25; a clean sheet in the first match means breaking the tight defense remains uncertain |
| Belgium vs Iran | Belgium has a slight edge, predicted score 1-0 | Belgium win 56.5% / Draw 23.3% / Iran win 20.1% | Belgium’s midfield and attacking have the upper quality and rating, expected goals around 1.79-0.83; Iran’s attack showed in the first match, with some surprise paths on counterattacks and set pieces |
| Uruguay vs Cape Verde | Uruguay has a slight edge, predicted score 2-0 | Uruguay win 68.1% / Draw 19.3% / Cape Verde win 12.6% | Uruguay has stronger physicality, set pieces, and attacking up front, expected goals around 2.08-0.53; Cape Verde swept Spain in their first match, and low post defense will increase the likelihood of a draw |
| New Zealand vs Egypt | Egypt slightly leads, predicted score 0-1 | New Zealand win 15.7% / Draw 21.3% / Egypt win 63.0% | Egypt overall rating and individual attacking ability are favored, expected goals around 0.67-1.95; New Zealand scored 2 goals in their opening match, with direct play and set pieces still able to create some fluctuations |
Warehouse Structure
worldcup-2026-predictor/ is the portable skill directory. You can copy it to a compatible AI skill runtime environment or run the scripts directly on the command line.
Installation
Requires Python 3.9+, scripts do not rely on third-party libraries. During installation, only copy the worldcup-2026-predictor/ directory; data/live/ is the runtime cache, no commit or distribution required.
Clone warehouse first:
Codex
Recommended installation is user-level skills, available for all Codex projects:
You can also install it as a project-level skill, copying the directory to the current project’s '.
This page was translated automatically. Technical details may differ from the original.
Recommended Tools
Try a different keyword or remove a filter.
Install
npx skillfish add ggen0488-cyber/worldcup-2026-predictor-skill